@WG-prioritization/alerts issue #81051 has been requested for prioritization.
I-nominated
?Apparently Windows 7 has been EOL for a year, so this is probably P-low if anything.
Yes though for the std lib Windows 7 is still supported by the Rust project. Dropping support for XP was only recently accepted Windows 7 is still supported. I don't think there's a solid policy on what supporting a target means. Do we support building on Windows 7? Only cross compiling to it?
This seems like something that should be clarified at some point
Not only is Windows 7 supported, apparently it's Tier 1.
For now i686-pc-windows-gnu
and x86_64-pc-windows-msvc
are tier 1 and the issue at hand is from a user that cannot compile on both, so we should uphold our promise of Tier 1 support.
If Win7 should be demoted because EOL'ed, it's fine but there's should be an agreement. Sending different messages to users is confusing.
All this said a permission denied
error doesnt sound like a Rust compiler issue :-) so for the moment I'd say P-low
until we understand more
by the way, I wonder if those disk volumes (F:
and G:
) that reporters are compiling onto are local or remote ... or even USB stuff
self-note: leaving this unprioritized, waiting for more info to clarify
removing the I-prioritize label and adding that we are waiting for a feedback form issue reporter
Issue #81051's prioritization request has been removed.
@WG-prioritization/alerts issue #81051 has been requested for prioritization.
I-nominated
?This makes rustc effectively unusable on windows 7
Sounds kind of bad :P
Pinged the LLVM group
well done @Léo Lanteri Thauvin . This issue now seems clearer thanks to the debugging done. I'd assign a P-high
hopefully someone in the team can comment on how easily this can be reproducible, as this seems affecting Win7 and the CI fails to catch it
I'd also like a more general question about Win7 support to be also touched (like also Camelid commented)
which to be clear could be also only tangent to this issue, though I don't understand the implication (why only on Win7 x64 and that LLVM commit?)
Issue #81051's prioritization request has been removed.
@WG-prioritization I wonder if this issue should be a regression-from-stable-to-nightly
. Since it's a) unassigned so I'm afraid it would pass unnoticed and b) going back to a previous nightly fixes the problem
Which version is nightly-2021-01-12?
(checking...)
Right, it's 1.51.0, so nightly soon-to-be-beta
(Applied the label)
ah thanks! how did you figure that out?
I just lazily executed rustup toolchain install nightly-2021-01-12
:
nightly-2021-01-12-x86_64-apple-darwin installed - rustc 1.51.0-nightly (c5eae5629 2021-01-11)
Though there must be a better way :D
Just looking at the date might be enough
I agree that if Windows 7 is a tier 1 platform; until that’s changed, we should treat it as so, and therefore treat this as a regression.
Though are we the ones who Tag regressions?
I think we do (when we have enough info to assess that, such as in this case)